> [!CAUTION]
> **Caution:** Support for backing up and restoring workloads managed using the legacy appliance management console is deprecated. For more information, see [Feature deprecations](https://docs.cloud.google.com/backup-disaster-recovery/docs/deprecations.md.txt) in the Backup and DR documentation.

This page includes information on how to ensure there is NFS connectivity for
Backup and DR.

## Install the NFS client on a Red Hat RHEL 6 or CentOS Linux host

See if the client is installed by running: `# rpm -qa | grep nfs`

This should return something like:

```bash
  nfs-utils-lib-1.1.5-9.el6.x86_64
  nfs-utils-1.2.3-54.el6.x86_64
```

- If you see nothing, then use yum to install the NFS client packages. Run: `# yum install nfs-utils nfs-utils-lib`
- Make sure rpcbind /portmapper package is installed on the Linux host. Run:`# rpm -qa | grep rpcbind`

This should return something like: `rpcbind-0.2.0-11.el6.x86_64`

- If you see nothing, then use yum to install the rpcbind. Run: `# yum install rpcbind`

## Install the NFS client on a SLES Linux host

1. To see if the client is installed, run: `# rpm -qa | grep nfs`

   This should return something similar to:

   ```bash
   nfs-client-1.2.1-2.6.6
   yast2-nfs-common-2.17.7-1.1.2
   yast2-nfs-client-2.17.12-0.1.81
   ```
2. If you don't see either nfs-client or yast2-nfs-xxxx packages, then use
   either YaST or zypper to install the NFS client packages.

   - Using YaST:

       # yast2 --install yast2-nfs-client
       # yast2 --install yast2-nfs-common

   - Using Zypper:`# zypper install nfs-client`
3. Ensure that rpcbind or portmapper package is installed on the Linux host. Run:
   `# rpm -qa | grep rpcbind`
   This should return something like: rpcbind-0.1.6+git20080930-6.15

4. If you see nothing, then you must install the packages using either YaST or
   zypper:

   - Using YaST: `# yast2 --install rpcbind`
   - Using Zypper: `# zypper install rpcbind`

## Learn NFS client information from the Linux host

A Backup and DR-approved NFS client package and version must be installed on
the host.

1. Check if the portmapper or rpcbind service is running. Run:
   `# sudo service rpcbind status`

   - A Red Hat RHEL 6 or CentOS Linux host should return something like: `rpcbind (pid 1591) is running...`
   - An SLES Linux host should return something like: `Checking for service rpcbind running`
2. If rpcbind service is not running on Linux host, start it with:
   `# sudo service rpcbind start`

3. Use rpcinfo to list the registered RPC programs or services. Portmapper
   must be registered and running.

       # sudo rpcinfo -p
       program vers proto port service
       100000 4 tcp 111 portmapper
       100000 3 tcp 111 portmapper
       100000 2 tcp 111 portmapper
       100000 4 udp 111 portmapper
       100000 3 udp 111 portmapper
       100000 2 udp 111 portmapper

4. Check if the Linux host can make an RPC call to rpcbind and NFS programs on
   the backup/recovery appliance using the following.

       # sudo rpcinfo -T tcp <#vm internal IP> rpcbind
       program 100000 version 2 ready and waiting
       program 100000 version 3 ready and waiting
       program 100000 version 4 ready and waiting
       # sudo rpcinfo -T tcp <#vm internal IP> nfs
       program 100003 version 2 ready and waiting
       program 100003 version 3 ready and waiting

If the preceding commands return the output shown earlier, then NFS connectivity from
Linux host to the backup/recovery appliance is good.

## Planning staging disk size

For some large filesystems, you may have to manually set the staging disk size
for the file system. The default staging disk size is `NAS capacity + 20%`,
but there are two cases in which this may be insufficient:

- NFS and SMB network file systems sometimes incorrectly report very large
  capacities. In cases where the file system reports that it is over 128TiB the
  Backup and DR agent fails the backup with error code 5289:
  "The reported size of the protected volume requires that the staging disk size
  is specified for this application". This error prevents Backup and DR
  from allocating a huge disk that is not needed or that is larger than the
  backup/recovery appliance can handle.

- Even if your NAS is using dedup and compression on its disks,
  Backup and DR does not deduplicate or compress data in the backup on
  the staging disk. Your NAS may report usage of 5TB, but the backup on
  the staging disk may use significantly more space. This also requires that the
  administrators specify a manual staging disk size. This may result in a
  "staging disk full" error.

If you see either of these errors, then manually set the staging disk size in
[Application Details and Settings](https://docs.cloud.google.com/backup-disaster-recovery/management-console/backup/app-details-settings-filesystems.md.txt).

## Virtual snapshots in a .snapshot directory

Sometimes on the NAS there are .snapshot directories containing a full copy of
the NAS contents. These are virtual snapshots of the NAS. The Backup and DR agent
tries to copy all of those snapshots and runs out of space. You can remedy this
by using an exclude pattern of `.snapshot` or `~snapshot` (whatever name the NAS
uses). See [Exclude patterns in](https://docs.cloud.google.com/backup-disaster-recovery/management-console/backup/app-details-settings-filesystems.md.txt).

## Additional information for preparing file system hosts

Additional information relevant to preparing a file system host for protection
are in [Manage hosts and their connected applications](https://docs.cloud.google.com/backup-disaster-recovery/management-console/configuration/manage-hosts-and-their-applications.md.txt).